Current Neurovascular Research, 2010, 7, 00-00
1
Traumatic Spinal Cord Injury Alters Angiogenic Factors and TGF-Beta1 that may Affect Vascular Recovery Marie-Françoise Ritz*,1, Ursula Graumann2, Bertha Gutierrez2 and Oliver Hausmann2,3 1
Department of Biomedicine, Brain Tumor Biology, Pharmacenter, Klingelbergstrasse 50, 4056 Basel, Switzerland
2
Swiss Paraplegic Research, Spinal Injury Research, Guido A. Zäch Strasse 4, 6207 Nottwil, Switzerland
3
Neurocenter Hirslanden Clinic St. Anna, St. Anna-Strasse 32, 6006 Lucerne, Switzerland Abstract: Traumatic spinal cord injury (SCI) disrupts the blood-spinal cord barrier and reduces the blood supply caused by microvascular changes. Vessel regression and neovascularization have been observed in the course of secondary injury contributing to microvascular remodeling after trauma. Spatio-temporal distribution of blood vessels and modulation of gene expression of several angiogenic factors have been investigated in rats after spinal cord compression injury. Rarefaction of vessels was detectable at the injury site 2 days after SCI before they disappeared in the developing cavity after 2 and 4 weeks, whereas no changes were observed in the penumbra. Investigation of the temporal expression of angiogenic genes using quantitative RT-PCR disclosed a constant down-regulation of the vascular endothelial growth factor (VEGF), and transient decreases of angiopoietin-1 (Ang-1), platelet-derived growth factor-BB (PDGF-BB), as well as placental growth factor (PlGF), with the lowest values obtained 3 days after injury, when compared to the expression levels obtained in sham-operated rats. Hepatocyte growth factor (HGF) was the only angiogenic factor with a constant increased gene expression when compared with controls, starting at day 3 post-SCI. mRNA levels of transforming growth factor-beta 1 (TGF-β1) were elevated at every time points following SCI, whereas those encoding for the cysteine-rich protein CCN1/CYR61 were upregulated after 2 h, 6 h, and 1 week only. Our data provide an overview of the temporal modulated expression of the major angiogenic factors, hampering revascularization in the lesion during the phase of secondary injury. These findings should be considered in order to improve therapeutic interventions.
Keywords: Angiogenic factors, gene expression, rat model, revascularization, spinal cord injury, vessel regression. INTRODUCTION Traumatic spinal cord injury (SCI) causes a primary injury followed by an induction of a cascade of secondary processes that collectively lead to additional tissue loss. Inflammatory processes as well as the breakdown of the blood-spinal cord barrier (BSCB) are described as the main causes for the secondary loss of neurons and glial cells, thus causing further deterioration of the patient. Rapid changes in the structural and biochemical integrity of spinal cord microvascular endothelia lead to a profound reduction of blood supply to the spinal cord. These changes affect neural and vascular cells likewise since they act in a close relationship as components of the neurovascular unit (reviewed in [1]). The regenerative capacity of the nervous tissue vasculature is well accepted, but it has been observed that newly formed vessels differ in morphology and show increased instability, ascribed to the missing association with astrocytes, decreased numbers of pericytes or to the deregulation of extracellular matrix [2-5]. Moreover, several studies have shown a biphasic response of the vascular system to injury of the rat spinal cord [5-7]. Indeed, an initial decrease of vessel density was observed at the lesion site, followed first by an induction of neovascularization, and then by a *Address correspondence to this author at the Department of Biomedicine, Brain Tumor Biology, Pharmacenter, Klingelbergstrasse 50, 4056 Basel, Switzerland; Tel: +41 61 267 15 35; Fax: +41 61 267 16 28; E-mail:
[email protected] Received: July 07, 2010
Revised: August 19, 2010
Accepted: September 02, 2010
1567-2026/10 $55.00+.00
regression coinciding with the development of the cavity [4]. However, the mechanisms triggering the decrease in vessel integrity are not fully understood. Recently, we highlighted the response of resident pericytes to SCI, demonstrating the disappearance of a CD133+ pericyte subpopulation in the early phase of secondary injury followed by their subsequent reappearance [8]. Mural cells are a requisite for the stabilization of newly formed and already established vessels, their decline will lead to endothelial cell (EC) apoptosis and consequently to vessel regression [2]. Association of pericytes and vascular smooth muscle cells (VSMCs) with newly formed vessels, as established at capillaries and arterioles, regulates EC proliferation, survival, migration, differentiation, morphogenesis, blood flow, and vessel permeability (reviewed in [9]). Therefore, not only EC growth and survival but also pericyte recruitment might play a crucial role during the microvascular remodeling in the secondary injury phase after SCI. Neovascularization is promoted by vasculogenesis and angiogenesis during wound healing after trauma [10,11]. While angiogenesis is defined as vascular sprouting from pre-existing blood vessels, the term vasculogenesis is equivalent with the process of neovascularization implemented by differentiating endothelial progenitor cells. However, the two models of neovascularization have in common the need of angiogenic factors as well as stabilizing agents for the durable establishment of functional blood vessels. Angiogenesis is fundamental in the development and in the adulthood, requiring the coordinated action of a variety of growth factors and cell-adhesion molecules. In the central nervous © 2010 Bentham Science Publishers Ltd.
2 Current Neurovascular Research, 2010, Vol. 7, No. 4
system (CNS) the situation is even more complex, since astrocytes and neurons are known to influence the endothelial phenotype [12]. Additionally, the close astrocyte-endothelial cell relationship is a major component of the BSCB integrity [4,13]. The contribution of angiogenic and counter-regulatory anti-angiogenic factors in the complicated course of vessel remodeling after SCI is highlighted by several reports [1417]. However, only few data exist so far regarding the role of anti-angiogenic factors in this process [14,18]. The best characterized angiogenic factors for vascular sprouting and stabilization are vascular endothelial growth factor A (VEGF-A), angiopoietin-1 (Ang-1), and plateletderived growth factor-BB (PDGF-BB). VEGF is the main inducer of angiogenesis [19] by initiating vessel formation and by enlarging small vessels [20]. It promotes formation of leaky, immature, and unstable vessels on its own. Such destabilized vessels would be prone to regression in the absence of associated growth factors. In contrast, Ang-1 induces vessel sprouting and branching [21,22], and seemingly further stabilizes and protects the adult vasculature, making it resistant to damage and leak by VEGF or inflammatory challenges [23]. PDGF-BB promotes neovascularization by recruiting pericytes [24] and priming VSMCs/pericytes to release pro-angiogenic mediators [25]. Several additional angiogenic factors are known to contribute to vessel growth and stabilization. The hepatocyte growth factor (HGF) is a potent angiogenic factor known to induce EC proliferation and migration [26], to ameliorate endothelial function, and to inhibit apoptosis [27]. The placental growth factor (PlGF) contributes to arteriogenesis and stabilization of vessels as documented by the recruitment of VSMCs under pathological conditions [28]. Transforming growth factor-beta 1 (TGF-β1) exhibits both pro- and antiangiogenic properties depending on the context and concentration [28]. The cysteine-rich angiogenic protein 61 CCN1/ CYR61 mediates adhesion, migration, and proliferation of vascular cells [29]. However, excessive accumulation of CCN1/CYR61 may initiate de-adhesion between pericytes and their surrounding ECM protein substrate [30]. For the development of new treatments to promote posttraumatic angiogenesis in the injured spinal cord a better understanding of the underlying mechanisms leading to vascular regression and failure of revascularization is needed. Several experimental attempts have been conducted to improve vessel growth in the CNS after injury by administration of exogenous angiogenic factors. Acute injection of VEGF alone after SCI was either neuroprotective [16] or exacerbated tissue loss with no vascular improvement [31]. In contrast, the treatment with Ang-1 and VEGF in combination induced a synergistic angiogenic effect, and promoted the formation of mature neovessels without the side effects of VEGF on the blood-brain barrier (BBB) in models of cerebral ischemia [32,33]. The aim of this study was to elucidate the endogenous angiogenic response in the rat spinal cord following a mild compression injury. Changes in the vascular compartment after SCI were analyzed by immunohistochemistry. The precise time-dependent expression profiles of VEGF, Ang-1, PDGF-BB, HGF, PlGF, TGF-β1, and CCN1/CYR61 were investigated by quantitative RT-PCR. Our results provide
Ritz et al.
new insight in the complex molecular reaction resulting in microvascular plasticity and dysfunction during the phase of secondary injury. MATERIAL AND METHODS Animals and Surgical Procedures All animal experiments were approved by the local Animal Welfare Committee and were in compliance with the Swiss Guidelines for the Care and Use of Animals. For SCI, adult male Wistar rats (250-300 g, Biological Research Laboratories Ltd, Füllinsdorf, Switzerland) were anesthetized through a nose mask with 3% isoflurane (Baxter AG, Volketswil, Switzerland) in a gas mixture of 30% oxygen and 70% nitrogen. Body temperature was maintained at 37°C using a rectal probe coupled with a homeothermic blanket system (Harvard Apparatus, Holliston, MA, USA) during anesthesia. Following laminectomy, the spinal cord at thoracic levels T8-T9 was compressed using a microvascular clip with a closing force of 20 g (Biemer, Mediwar AG, Switzerland) for 15 sec. Controls consisted of sham animals that received a laminectomy but were not subjected to spinal cord injury. Subsequently, the incision was closed in two layers and the animals were allowed to recover from anesthesia and returned to their home cage. After survival times of 2, 6, 24, 48, 72 hours, 1, 2, 3, and 4 weeks, animals were deeply anesthetized with isoflurane, decapitated, and spinal cords were immediately dissected. For gene expression analysis, segments of 5 mm spinal cord containing the compression site were fresh frozen on dry ice before further processing. For the time points chosen for immunohistology, spinal cord tissue segments of 10 mm containing the injury site were immediately embedded in O.C.T. (Sakura, NL), and frozen at -80°C. Immunohistochemistry Horizontal spinal cord sections (10 µm) from shamoperated animals (n=3) and injured animals after 2 days, 2 weeks, and 4 weeks following SCI (n=3 for each time point) were cut in a cryostat and mounted on Super Frost Plus slides (Menzel, Braunschweig, Germany). After fixation with acetone at 4°C for 10 min, sections were washed in PBS, incubated with blocking buffer (20% normal goat or donkey serum in PBS containing 0.2% Triton X-100 (PBST)) for 20 min and then incubated overnight at 4°C with rat endothelial antibody-1 (RECA-1) raised in mouse from Abcam (ab22492, UK) diluted 1/250. Subsequently, sections were washed with PBS, and incubated for 30 min at room temperature with Cy3-coupled secondary antibodies. For some sections, co-staining with DAPI (4’,6’-diamidine-2phenylindole) was performed in order to visualize the cell nuclei. Sections were washed in PBS and covered with Fluorosave (Calbiochem). Sections were observed and photographed using a Leica DMRE microscope equipped with a F-view camera (Soft Imaging System, Germany). Immunohistological Quantification Quantifications of the RECA-1 fluorescent signal and of the vessel density were performed in regions of interest (ROI) in normal appearing spinal tissue adjacent to the
Angiogenic Response after Spinal Cord Injury
Table 1.
Current Neurovascular Research, 2010, Vol. 7 No. 4
3
List of Designed Oligonucleotide Primers Used for qRT-PCR
Gene
GeneBank
Forward
Reverse
vWF
XM_342759
5’-CCTCAGCACTGTCAGAATTGTC-3’
5’-GGGTGTCCTCAACATATGG-3’
HGF
NM_017017
5’-GGACCTTGTGAGGGAGATTATG-3’
5’-TACCAGGACGATTTGGGATG-3’
CCN1/CYR61
NM_031327
5’-GTGCCGCCTGGTGAAAGAGA-3’
5’-GCTGCATTTCTTGCCCTTTTTTAG-3’
TGF-β1
NM_021578
5’-CAATTCCTGGCGTTACCTTG-3’
5’-AAAGCCCTGTATTCCGTCTC-3’
VEGF-A
NM_031836
5’-GAGTCTGTGCTCTGGGATTTG-3’
5’-TCCTGCTACCTCTTTCCTCTG-3’
PDGF-BB
NM_031524
5’-AGCCAAGACACCTCAAACTC-3’
5’-AAATAACCCTGCCCACACTC-3’
60S/L13A
NM_173340
5’-ATCCCTCCACCCTATGAC-3’
5’-GTCACTGCCTGGTACTTC-3’
lesion sites (n=3 rats for each time point). The images of the ROI (3.9 mm2, 25X magnification for signal quantification, 100X magnification for the measurement of vessel densities) were binarized, analyzed with the AnalySIS Image program (AnalySIS, Soft Imaging System), and expressed as percentages of the ROI covered by the RECA-1 signal (% of ROI) or number of vessels per ROI (density of RECA-1 immunoreactive vessels). Gene Expression Analysis Total RNA was extracted from injured rats after post-SCI times of 2, 6, 24, 48, 72 hours, 1, 2, 3, and 4 weeks (n=3 for
all time points) and sham-operated rats (n=3 for all time points) using the RNeasy Lipid Tissue Kit (Qiagen, Hombrechtikon, Switzerland) including homogenization of the tissues with the Dispomix® Drive (Medic Tools AG, Zug, Switzerland). Following analysis of the RNA integrity by Experion Electrophoresis (Bio-Rad Laboratories AG, Reinach, Switzerland), 1 µg of RNA was transcribed into cDNA with Random Hexamers using First Strand cDNA Synthesis Kit for RT-PCR (Roche, Rotkreuz, Switzerland). For the absolute quantification of mRNA, gene-specific primers for von Willebrand Factor (vWF), HGF, TGF-β1, VEGF-A, PDGF-BB, and 60S ribosomal protein L13A
Fig. (1). Compression-induced changes in the spinal cord vasculature. A. Representative images of spinal cord sections stained with the endothelial cells marker RECA-1 from sham-operated rats (a), and operated rats after 2 days (b), 2 weeks (c), and 4 weeks (d) post-SCI. The centers of the compression sites are marked with an asterisk (), and the developing cavities (in c and d) are delineated. Scale bar=500 µm. B. The fraction of the area occupied by RECA-1 immunoreactive vessels was compared between sham-operated controls and the three time points (2 days, 2 and 4 weeks) in the normal-appearing parenchyma. *: P
